File: P o r t er ( f in a l) . d o c
C r ea t ed o n : 1 2 / 2 2 / 2 0 0 6 9 :0 5 :0 0 P M
L a s t P r in t ed : 1 2 / 2 3 / 2 0 0 6 1 :1 1 :0 0 P M
The Potential and Peril of BAPCPA for
E m p iric al R es earc h
Katherine Porter
*
I. IN T R O D U C T I O N .............................................................................................. 9
II. HI S T O R Y O F BA N K R U P T C Y DA T A ................................................................. 9
III. DA T A AN D RE S E A R C H BE N E F I T S O F BAPCPA ........................................... 9
A. R aw D ata S ou rc es ................................................................................ 9
B . G ov ernm ent R es earc h S tu d ies .............................................................. 9
C . F u tu re Ap p roac h to S tatis tic s ............................................................... 9
IV . PI T F A L L S I N RE S E A R C H AF T E R BAPCPA................................................... 9
A. D ata G ap s ............................................................................................ 9
B . R es earc h O b s tac l es .............................................................................. 9
V . CO N C L U S I O N ................................................................................................ 9
6 3
6 5
7 0
7 0
7 2
7 6
7 8
7 8
8 0
8 2
I. INTRODUCTION
Lamentations and hand-wringing accompanied the passage of the B ankruptcy A buse P revention and Consumer P rotection A ct (B A P CP A ).1 The last
y ear has shown that the amendments have indeed created troubling issues.2
* Associate Professor, University of Iow a College of Law . I th ank Mich elle
Arnopol Cecil for inviting me to participate in th e sy mposium. Jennifer K napp and
Ariane Holtsch lag assisted w ith my research .
1. Bankruptcy Abuse Prevention and Consumer Protection Act of 20 0 5, Pub. L.
No. 10 9-8, 119 Stat. 23, [ h ereinafter BAPCPA] (All furth er references to sections of
BAPCPA th at w ill be codified are to th e United States Code. References to oth er
sections of BAPCPA w ill be to th e public law sections) . Numerous professors predicted th at th e bill w ould h arm th e bankruptcy sy stem. See, e.g., Ch arles Tabb, Th e
D eath o f C o nsumer B ankruptc y in th e U nited States? , 18 BANKR. DEV . J. 1 (20 0 1) .
Th e American Bar Association also lobbied against th e bill, asserting th at attorney
liability provisions w ould h ave “ disastrous” conseq uences for th e entire bankruptcy
court sy stem. See Letter from th e Am. Bar Ass’n to Sen. Frist and Sen.
Dasch le (Nov. 21, 20 0 3) , av ailable at h ttp:/ / w w w .abanet.org/ poladv/ letters/ 10 8th / bankruptcy 11210 3.pdf.
2. Commentary on a variety of problems w ith BAPCPA, including cases struggling to interpret th e new law , is available at a current blog. See ABI’s BAPCPA
Blog, h ttp:/ / bapcpa.blogspot.com (last visited July 18, 20 0 6) ; see also Trying to M ake
Sense o ut o f N o nsense: R epresenting C o nsumers U nd er th e “ B ankruptc y A buse P rev entio n and C o nsumer P ro tec tio n A c t o f 2 0 0 5 ,” 79 AM. BANKR. L.J. 191 (20 0 5) .
Courts are struggling w ith BAPCPA’s provisions as w ell. See, e.g., I n re K ane, 336
B.R. 477 (Bankr. D. Nev. 20 0 6) (addressing purported drafting mistake in BAPCPA
provision) . Th e first large empirical study of post-BAPCPA cases reported problems
w ith th e bill’s credit counseling provisions. See NAT’L ASSN. OF CONSU MER BANKR.
File: P o r t er ( f in a l) . d o c
9 6 4
C r ea t ed o n : 1 2 / 2 2 / 2 0 0 6 9 :0 5 :0 0 P M
M I S S O U R I L AW
R E V IE W
L a s t P r in t ed : 1 2 / 2 3 / 2 0 0 6 1 :1 1 :0 0 P M
[ V o l.7 1
This sy mposium identifies ambiguities and problems in the amended B ankruptcy Code and begins to untangle the effects of the new law on the bankruptcy sy stem.3 These provocative issues are fundamentally research q uestions. The next few y ears are likely to powerfully demonstrate how comprehensive law reform such as B A P CP A whets the academic appetite for research.
B A P CP A does not j ust raise research q uestions, however. The law itself
actually seeks to provide answers. The statute has numerous explicit provisions about statistics and studies,4 and the law’s emphasis on disclosure implicitly increases the amount of data about people and businesses in financial
trouble.5 The central place of empirical research in B A P CP A reflects the importance of data in modern policy making. Indeed, much of the bankruptcy
reform debate was a battle of numbers.6 B y including research mandates in
the new law, Congress articulated an empirical research agenda about bankruptcy for the federal government. This intertwining of research and legal
reform provides an unprecedented moment to advance bankruptcy policy making based on empirical reality , rather than theoretical speculation.
This article survey s the history of bankruptcy data and identifies the
B A P CP A provisions that bear directly on research. It concludes by examining
how such studies will and should proceed. B A P CP A provides both opportunities and haz ards to advance our understanding of bankruptcy . The development of comprehensive federal data offers the potential to dramatically increase the scope of knowledge about the bankruptcy sy stem. The peril lies in
the government conducting its research without the transparency and accountability necessary to convince private industry , academic scholars, and
the general public of the integrity and usefulness of these data. R ather than
eclipsing academic research, the federal government’s bold new foray into
empirical bankruptcy work challenges the scholarly community to engage
ATTORNEY S, BANKRU P TC Y REF ORM’S IMP AC T: WH ERE ARE AL L TH E “ DEADB EATS? ”
(Feb.
22,
20 0 6)
[ h ereinafter
DEADB EATS],
av ailable
at
h ttp:/ / nacba.com/ files/ main_ page/ 0 2220 6NACBAbankruptcy reformstudy .pdf (concluding th at few consumers can use debt management programs or can pay any th ing
on th eir debts but th at credit counseling is an additional burden before bankruptcy ) .
3. Some of th e pieces in th is sy mposium volume comment on th e process th at
sh aped th e legislation, and discuss w h y BAPCPA failed to address certain areas for
reform. Professor K eating and Professor Dickerson offer insigh ts on th is issue. Oth er
articles, such as th ose contributed by Professor Jacoby and Judge Wedoff, speculate
about h ow BAPCPA w ill affect th e problems th at people bring to bankruptcy and h ow
j udges sh ould apply th e law to th ese problems.
4. See inf ra Part III.B.
5. See inf ra Part III.A.
6. See, e.g., Eliz abeth Warren, Th e P h anto m $ 4 0 0 , 13 J. BANKR. L. & PRAC . 77
(20 0 4) ; Eliz abeth Warren, Th e M arket f o r D ata: Th e C h angin R o le o f So c ial Sc ienc es
in Sh aping th e L aw , 20 0 2 WIS. L. REV . 1, 13-16; Marianne B. Culh ane & Mich aela
M. Wh ite, D ebt A f ter D isc h arge: A n Empiric al Stud y o f R eaf f irmatio n, 73 AM.
BANKR. L.J. 70 9 (1999) .
File: P o r t er ( f in a l) . d o c
2 0 0 6 ]
C r ea t ed o n : 1 2 / 2 2 / 2 0 0 6 9 :0 5 :0 0 P M
R E S E AR C H I N G B APC PA
L a s t P r in t ed : 1 2 / 2 3 / 2 0 0 6 1 :1 1 :0 0 P M
9 6 5
with government and private industry to ensure collective improvement of
bankruptcy knowledge. If these collaborations succeed, the result will be a
new world of research that features more reliable empirical data and facilitates a better understanding of the bankruptcy sy stem.
II. HISTORY OF BANK
RUPTCY DATA
E mpirical research about bankruptcy is relatively y oung.7 A cademics
such as M arj orie G irth, P hillips S huchman, and the trio of S ullivan, W arren,
and W estbrook pioneered empirical research about bankruptcy . This early
work was crucial to sparking interest in the bankruptcy sy stem and helped to
shape the development and interpretation of the 1 9 7 8 B ankruptcy Code.8
D uring subseq uent rounds of bankruptcy law-making, the credit industry
began to use data to support its positions.9 Industry -funded studies actively
competed with independent academic research in what E liz abeth W arren has
termed a “new marketplace for empirical data.”10 The number of female
bankruptcy filers, the amount of consumer debt discharged in bankruptcy
each y ear, and the percentage of small businesses in bankruptcy all were crucial and contested facts in the bankruptcy reform debates.
The federal government has been a reluctant participant in empirical research about bankruptcy . In fact, most federal studies ignore bankruptcy entirely . The U .S . Census B ureau does not collect information about bankruptcy
in its survey s, despite the fact that filing bankruptcy is as common of an experience as getting divorced or earning a college degree,11 both of which are
captured by Census data. The S mall B usiness A dministration, the D epartment
of D efense’s military personnel studies, and the S urvey of Consumer F inance
all either ignore bankruptcy or limit themselves to asking if a respondent has
ever filed. Calls for improved information about the bankruptcy sy stem date
at least to 1 9 7 3 .12
7. Th e first large empirical study of bankruptcy w as conducted in 1964 and
funded by th e Brookings Institute. See DAV ID T. STANL EY & MARJ ORIE GIRTH ,
BANKRU P TC Y : PROB L EM, PROC ESS & REF ORM (1971) . Prior studies w ere small and
examined only one bankruptcy district. See TERESA A. SU L L IV AN, EL IZ AB ETH
WARREN & JAY LAW RENC E WESTB ROOK, FRAG IL E MIDDL E CL ASS 34, 296 (20 0 0 )
(summariz ing th ese studies) .
8. See Bruce G. Carruth ers & Terence C. Halliday , P ro f essio nals in Systemic
R ef o rm o f B ankruptc y L aw : Th e 1 9 78 U .S. B ankruptc y C o d e and th e English I nso lv enc y A c t 1 9 8 6 , 74 AM. BANKR. L.J. 35 (20 0 0 ) .
9. See Warren, Th e M arket f o r D ata, supra note 6, at 12-14 (describing credit
industry studies during bankruptcy reform debates) .
10 . I d . at 39.
11. Eliz abeth Warren, W h at I s a W o men’ s I ssue? B ankruptc y and C o mmerc ial
L aw and O th er G end er-N eutral To pic s, 25 HARV . WOMEN’S L.J. 19, 21 (20 0 1) .
12. NAT’L BANKR. REV IEW COMM’N, BANKRU P TC Y : TH E NEX T TW ENTY YEARS
923 & n.2393 (1997) [ h ereinafter BANKRU P TC Y ], av ailable at
File: P o r t er ( f in a l) . d o c
9 6 6
C r ea t ed o n : 1 2 / 2 2 / 2 0 0 6 9 :0 5 :0 0 P M
M I S S O U R I L AW
R E V IE W
L a s t P r in t ed : 1 2 / 2 3 / 2 0 0 6 1 :1 1 :0 0 P M
[ V o l.7 1
S ince the enactment of the B ankruptcy Code, academics have outstripped the government in cultivating bankruptcy data. P erhaps the most
prominent example of this work is the ongoing Consumer B ankruptcy P roj ect, which presently consists of three large general studies of consumer bankruptcy conducted by D r. Teresa S ullivan, E liz abeth W arren, J ay W estbrook
and other researchers.13 In recent y ears, there has been a substantial increase
in the number and specializ ation of empirical studies.14 M any of these academic research findings were contested.15 F or example, R obert R asmussen
recently authored Empirically Bankrupt, which challenges the methodology
and conclusions of three empirical studies.16 W hile these critiq ues were not
alway s mild,17 the nature of academic research made these debates possible
because researchers generally disclosed their data to critics and made their
methodology explicit. This “open source”18 methodology is the hallmark of
academic empirical work.19 P rofessor Ly nn LoP ucki’s database of large, public company bankruptcies is the epitome of this norm about shared data in the
h ttp:/ / govinfo.library .unt.edu/ nbrc/ report/ 21bdata.pdf (citing REP ORT OF TH E COMLAW S OF TH E UNITED STATES, H.R. DOC . NO. 93-137,
pts. 1 & 2, at 111 (1973) ) .
13. For a description of th e 20 0 1 Consumer Bankruptcy Proj ect and a summary
of prior studies, see EL IZ AB ETH WARREN & AMEL IA WARREN TY AG I, TH E TW OINC OME TRAP 181-82 (20 0 3) .
14. See, e.g., Rafael I. Pardo & Mich elle R. Lacey , U nd ue H ard sh ip in th e B ankruptc y C o urts: A n Empiric al A ssessment o f th e D isc h arge o f Ed uc atio nal D ebt, 74 U.
CIN. L. REV . 40 5 (20 0 5) ; K ath erine Porter, G o ing B ro ke th e H ard W ay: Th e Ec o no mic s o f R ural F inanc ial F ailure, 20 0 5 WIS. L. REV . 969; Brian L. Betker, Steph en P.
Farris & Robert M. Law less, “ W arm w ith Sunny Skies”: D isc lo sure Statement F o rec asts, 73 AM. BANKR. L.J. 80 9 (1999) ; Scott F. Norberg, C o nsumer B ankruptc y’ s N ew
C lo th es: A n Empiric al Stud y o f D isc h arge and D ebt C o llec tio n in C h apter 1 3 , 7 AM.
BANKR. INST. L. REV . 415 (1999) ; Culh ane & Wh ite, supra note 6.
15. For example, Judge Edith Jones and Todd Z w y icki h ave criticiz ed data from
th e Consumer Bankruptcy Proj ects. See Warren, Th e M arket f o r D ata, supra note 6,
at 36-37 (documenting th e critiq ue) .
16. Robert K . Rasmussen, Empiric ally B ankrupt (V anderbilt Law
&
Econ.
Research
Paper
No.
0 6-0 7) ,
av ailable
at
h ttp:/ / papers.ssrn.com/ sol3/ papers.cfm? abstract_ id= 895547.
17. See, e.g., Eliz abeth Warren, Th e U ntenable C ase f o r R epeal o f C h apter 1 1 ,
10 2 YAL E L.J. 437, 453-54 (1992) (critiq uing Mich ael Bradley & Mich ael
Rosenz w eig, Th e U ntenable C ase f o r C h apter 1 1 , 10 1 YAL E L.J. 10 43 (1992) and
alleging a “ gross error” by Bradley & Rosenz w eig in analy z ing data) .
18. “ Open source” is a tech nology term th at describes softw are for w h ich th e
source code is made publicly available for replication or modification. For an open
source
definition
of
th e
term
open
source,
see
Wikipedia,
h ttp:/ / en.w ikipedia.org/ w iki/ Open_ source# _ note-0 (last visited Nov. 17, 20 0 6) .
19. Lee Epstein & Gary K ing, Th e R ules o f I nf erenc e, 69 U. CH I. L. REV . 1, 38
(20 0 2) (explaining th at replication is a basic test for evaluating th e q uality of empirical w ork) .
MISSION ON TH E BANKRU P TC Y
File: P o r t er ( f in a l) . d o c
2 0 0 6 ]
C r ea t ed o n : 1 2 / 2 2 / 2 0 0 6 9 :0 5 :0 0 P M
R E S E AR C H I N G B APC PA
L a s t P r in t ed : 1 2 / 2 3 / 2 0 0 6 1 :1 1 :0 0 P M
9 6 7
bankruptcy field.20 It is free, easily accessed, and makes its methodology
transparent. W hile concern about respecting the rights of human subj ects does
not alway s permit releasing all data, the academic community tends to treat
data as collective goods to contribute to shared knowledge.
The recent bankruptcy reforms demonstrated that academics no longer
enj oy a monopoly on empirical research about bankruptcy . The credit industry actively solicited Congress’s attention for private research that the industry had funded and had distributed selected findings to the media.21 F or example, the work of the Credit R esearch Center was a very high profile part of
the congressional debate about bankruptcy reform.22 Industry -funded studies
tried to measure how much bankruptcy debt is discharged and how much
creditors could recover if the law were changed.23 This private, industry funded research arguably threatens to eclipse the work of professional academics in influencing policy making.24
The government had little neutral data to use in settling these controversies. S ince 1 9 4 8 , the A dministrative O ffice of the U .S . Courts (“A O ”) has
been tasked with providing aggregate statistics on the bankruptcy court sy stem to the J udicial Conference.25 The A O releases the number of bankruptcy
filings broken down by chapter, business and non-business, and j udicial district to the public.26 These numbers are widely reported in the media, and
used by some analy sts as economic indicators.27 F or others, the trend of filings serves as a moral compass that indicates changing values about borrowing.28 Y et, these data do not reveal any thing about who uses the bankruptcy
20 . Professor LoPucki h as loftily named h is database th e “ Bankruptcy Research
Database.” Details about its scope and meth odology , as w ell as raw data, are available
at h ttp:/ / lopucki.law .ucla.edu (last visited July 18, 20 0 6) .
21. See Warren, Th e P h anto m $ 4 0 0 , supra note 6 (tracing th e h istory and use of
th e infamous $ 40 0 “ bankruptcy tax” figure) .
22. See Warren, Th e M arket f o r D ata, supra note 6, at 30 .
23. See K im J. K ow alew ski, Ev aluatio ns o f Th ree Stud ies Submitted to th e N atio nal B ankruptc y R ev iew C o mmissio n, in BANKRU P TC Y : TH E NEX T TW ENTY YEARS
G-2.d app. at 7-8 (1997) .
24. Warren, Th e M arket f o r D ata, supra note 6, at 43.
25. 28 U.S.C. § 60 4(a) (2) , (13) (1948) (current version at 28 U.S.C. § 60 4(a) (2) ,
(13) (20 0 0 ) .
26. In recent y ears, th e AO h as used its w ebsite to release th ese statistics. Th ey
h ave posted some prior data as w ell. See U.S. Courts: Bankruptcy Statistics,
h ttp:/ / w w w .uscourts.gov/ bnkrpcty stats/ bankruptcy stats.h tm (last visited Nov. 17,
20 0 6) .
27. See e.g., Alan Deaton, L arge and Small C o mpanies Ex h ibit D iv erging B ankruptc y
Trend s,
BANK
TRENDS
(Jan.
20 0 2) ,
av ailable
at
h ttp:/ / w w w .fdic.gov/ bank/ analy tical/ bank/ bt0 20 1.h tml; V ern McK inley , B allo o ning
B ankruptc ies (1997) , h ttp:/ / w w w .cato.org/ pubs/ regulation/ reg20 n4f.h tml (last visited
Nov. 17, 20 0 6) .
28. See e.g., Judge Edith H. Jones & Todd Z y w icki, I t’ s Time f o r M eans Testing,
1999 BYU L. REV . 177.
File: P o r t er ( f in a l) . d o c
9 6 8
C r ea t ed o n : 1 2 / 2 2 / 2 0 0 6 9 :0 5 :0 0 P M
M I S S O U R I L AW
R E V IE W
L a s t P r in t ed : 1 2 / 2 3 / 2 0 0 6 1 :1 1 :0 0 P M
[ V o l.7 1
sy stem, what problems they seek to solve, and whether the sy stem functions
effectively . Their main use is as time series data,29 and even in this regard,
they have been faulted for inaccuracy .30 The lack of transparency about how
these statistics are compiled spurs concerns about their reliability .31 W ithout
individual data, the “off-the-rack” government statistics do not permit meaningful analy sis. B are filing rates fail to reveal the subtleties of the bankruptcy
sy stem. They do not reveal how many debtors each y ear are y oung, female,
disabled, unemploy ed, divorced, self-employ ed, or subj ect to garnishment.
The official federal data gives the public only one gross figure to evaluate our
bankruptcy sy stem—the number of cases.32 A s the N ational B ankruptcy R eview Commission noted in its report on “D ata and D issemination,” the A O
“collects data for its own internal administrative purposes, not for the purpose
of monitoring the efficient operation of the B ankruptcy Code” and that “public interest data” are not regularly collected.33
W hile the A dministrative O ffice has only dabbled in data production,
the E xecutive O ffice of the U .S . Trustee (“U S T”), a branch of the D epartment
of J ustice,34 has engaged in some serious empirical research. F or y ears, E d
F ly nn of the U S T produced a column entitled “B ankruptcy by the N umbers”
for the A me rican Bankruptcy I ns titute J o urnal with co-author G ordon B ermant.35 These studies examined issues such as the number of elderly bankruptcy debtors, the j ob tenure of debtors, and the pay ment of mortgages
through Chapter 1 3 plans. F ly nn and B ermant generally documented their
methodology , which principally relied on public bankruptcy court records.
The U S T’s advantage in using these records is the cooperation of bankruptcy
29. See e.g., Robert M. Law less, Th e P arad o x o f C o nsumer C red it (Univ. Ill.
Law
&
Econ.
Research
Paper
No.
LE0 6-0 15) ,
av ailable
at
h ttp:/ / ssrn.com/ abstract= 90 6868 (comparing bankruptcy filing rate over time w ith
different measures of consumer credit) ; ROB ERT D. MANNING , CREDIT CARD NATION
127-29 fig.5.1 (20 0 0 ) (comparing trend of bankruptcy filing w ith unemploy ment rate
over time) .
30 . See, e.g., Eliz abeth Warren & Robert Law less, Th e M yth o f th e D isappearing
B usiness B ankruptc y, 93 CAL . L. REV . 743 (20 0 5) (critiq uing AO data) ; Jennifer Frasier, C augh t in a C yc le o f N eglec t: Th e A c c urac y o f B ankruptc y Statistic s, 10 1 COM.
L.J. 30 7 (1996) (documenting inconsistencies in AO data collection) .
31. BANKRU P TC Y , supra note 12, at 923, 926 (complaining about inconsistent
procedures in collecting data) .
32. I d . at 923 (“ About th e only data accurately reported are th e number of bankruptcy cases actually filed.” ) .
33. I d . at 928.
34. Th e U.S. Trustee Program is responsible for overseeing th e administration of
bankruptcy cases. See U.S. Trustee Program, h ttp:/ / w w w .usdoj .gov/ ust (last visited
Nov. 17, 20 0 6) .
35. Th ese reports and oth er articles by UST officials are available at
h ttp:/ / w w w .usdoj .gov/ ust/ eo/ public_ affairs/ articles.
File: P o r t er ( f in a l) . d o c
2 0 0 6 ]
C r ea t ed o n : 1 2 / 2 2 / 2 0 0 6 9 :0 5 :0 0 P M
R E S E AR C H I N G B APC PA
L a s t P r in t ed : 1 2 / 2 3 / 2 0 0 6 1 :1 1 :0 0 P M
9 6 9
court clerks in gathering district-level data.36 The U S T also enj oy s free access
to electronic records in the P ublic A ccess to Court E lectronic R ecords
(P A CE R ) sy stem.37 O thers must pay a per-page fee or obtain a fee waiver
from the chief j udge in each district in which they wish to view the records.
F or empirical research, these fees can be substantial, as is the time and energy
commitment req uired to follow through on persuading each j udge to grant a
fee waiver. These disadvantages hinder non-government researchers and may
prevent independent or academic scholars from replicating federal studies.
W hile most federal researchers have either ignored bankruptcy or
merely produced a few aggregate statistics, the G overnment A ccountability
O ffice did produce a few studies in the y ears preceding B A P CP A .38 O ne such
study critically evaluated the claims of the credit industry about how much
debtors can pay on their debts,39 and attempted to determine the relative reliability of this research.40 The government was neither a repository for much
important data about bankruptcy nor a producer of leading research; it merely
examined existing research. The N ational B ankruptcy R eview Commission
criticiz ed the “dearth of timely , accurate bankruptcy data” and proposed numerous improvements to the data collection sy stem.41
B A P CP A changes the status q uo in bankruptcy research. B A P CP A
mandates public research to be conducted by the G A O , the U .S . Trustee, and
other federal agencies,42 thus adding a third play er to the existing contest over
bankruptcy data. These provisions provide an opportunity to head off the
growing collision between scholars and for-hire researchers by creating
shared, reliable, neutral, and transparent data. W hether B A P CP A will fulfill
this potential is presently unclear. Its success will depend largely on the willingness of the government, private researchers, and academics to collaborate,
36. Academic research ers obtain th ese data, if at all, w ith w inning smiles, lots of
patience, and good luck.
37. Th e PACER sy stem is administered by th e Administrative Office of th e U.S.
Courts.
38. See, e.g., U.S. GEN. AC C OU NTING OF F IC E, BANKRU P TC Y REF ORM: USE OF
TH E HOMESTEAD EX EMP TION B Y
CH AP TER 7 BANKRU P TC Y DEB TORS IN TH E
NORTH ERN DISTRIC T OF TEX AS AND TH E SOU TH ERN DISTRIC T OF FL ORIDA, GGD-99142R (July 12, 1999) ; GEN. AC C OU NTING OF F IC E, BANKRU P TC Y ADMINISTRATION:
CASE REC EIP TS PAID TO CREDITORS AND PROF ESSIONAL S, GGD-94-173 (July 13,
1994) . “ GAO” is now th e acrony m for th e Government Accountability Office, w h ich
is th e new title for th e former General Accounting Office.
39. See supra notes 21-23 and accompany ing text.
40 . U.S. Gen. Accounting Office, P erso nal B ankruptc y A nalysis o f F o ur R epo rts
o n C h apter 7 D ebto rs’ A bility to P ay, GAO/ GGD-99-10 3, 1, 5 (June 1999) (explaining th at its analy sis w as h ampered by industry research ers refusal to release data) ;
U.S. Gen. Accounting Office, P erso nal B ankruptc y: Th e C red it R esearc h C enter
R epo rt o n D ebto rs’ A bility to P ay, GAO/ GGD-98-47, at 6 (Feb. 1998) .
41. BANKRU P TC Y , supra note 12, at 921-22.
42. See inf ra Part III.B. (detailing th ese BAPCPA provisions) .
File: P o r t er ( f in a l) . d o c
9 7 0
C r ea t ed o n : 1 2 / 2 2 / 2 0 0 6 9 :0 5 :0 0 P M
M I S S O U R I L AW
L a s t P r in t ed : 1 2 / 2 3 / 2 0 0 6 1 :1 1 :0 0 P M
R E V IE W
[ V o l.7 1
as well as the stimulation of a persistent, collective demand for accurate government bankruptcy data.
III. DATA AND RESEARCH BENEFITS OF BAPCPA
B A P CP A should enhance the available data about bankruptcy in two
way s. The reformed law makes implicit contributions to research by putting
new duties on debtors and by expanding the req uired bankruptcy form. Industrious researchers can painstakingly transform these enhanced records into
usable data. B A P CP A also contains explicit research mandates and req uires
multiple formal studies of certain aspects of the bankruptcy sy stem.
A. Raw Data Sources
D espite being much maligned by professors,43 the means test may have
a silver lining. The J udicial Conference A dvisory Committee on B ankruptcy
R ules has promulgated F orm B 2 2 A to implement the means test.44 To establish a debtor’s “current monthly income”45 for use in the means test, F orm
2 2 A req uires all Chapter 7 debtors to report their average income for the six
calendar months before filing bankruptcy . The form also req uires debtors to
break this income out by the source of the income. These are new data,46 and
they should be highly reliable given the additional req uirement of providing
pay ment advices47 and the sharp penalty for misreporting on bankruptcy
forms.48 These data allow comparisons between debtors’ past income and
what they are earning at the time of their bankruptcy filings. They could also
be useful to identify debtors who operated a business or received unemploy ment benefits before bankruptcy . A dditionally , these data could suggest what
ty pes of risks and problems precede financial failure.
43. Law Professors’ Letter to Senator Specter and Senator Leah y (Feb. 16. 20 0 5)
(critiq uing th e means test on numerous grounds) , reprinted in Ch arles Jordan Tabb,
C o nsumer B ankruptc y A f ter th e F all: U nited States L aw U nd er S. 2 5 6 , CANADIAN
BU S.
L.
J.
(forth coming) ,
av ailable
at
h ttp:/ / papers.ssrn.com/ sol3/ papers.cfm? abstract_ id= 859645.
44. Office Form B22A is available at th e U.S. Courts w ebpage, Bankruptcy
Forms, h ttp:/ / w w w .uscourts.gov/ bkforms/ index.h tml (last visited July 18, 20 0 6) .
45. 11 U.S.C. § 10 1(10 A) (Supp. V 20 0 5) .
46. Previously , th e only data on past income came from th e Statement of Financial Affairs, w h ich asked for annual income for each of th e th ree y ears before bankruptcy . From my experience coding bankruptcy court records as part of th e 20 0 1
Consumer Bankruptcy Proj ect, a large national empirical study , I can attest th at income fields on th e Statement of Financial Affairs w ere freq uently reported inconsistently , sometimes only partially completed, and occasionally even left entirely blank.
47. 11 U.S.C. § 521(a) (1) (B) (iv) .
48. I d . § 110 (20 0 0 & Supp. V 20 0 5) .
File: P o r t er ( f in a l) . d o c
2 0 0 6 ]
C r ea t ed o n : 1 2 / 2 2 / 2 0 0 6 9 :0 5 :0 0 P M
R E S E AR C H I N G B APC PA
L a s t P r in t ed : 1 2 / 2 3 / 2 0 0 6 1 :1 1 :0 0 P M
9 7 1
The expense information req uired by the means test will likely be of
limited value. O nly debtors whose current monthly income exceeds the applicable median family income are req uired to disclose expenses on F orm
B 2 2 A . F urther, the means test attributes fictional expenses to debtors based
on Internal R evenue S ervice standards.49 W hen actual expenses are reported,
however, such data could prove helpful in revealing the ty pes of pressures
that push families with substantial incomes to their financial breaking point.
F or example, research may show that these families freq uently have very high
expenses for health care, childcare, or domestic support obligations. I doubt
that this data will prove worth the burden that its collection imposes on debtors,50 but thoughtful scholars may be able to tease valuable insights from
them.
B A P CP A ’s amendments to Chapter 1 3 also req uire enhanced disclosures. F orm B 2 2 C req uires Chapter 1 3 debtors to make calculations of income and expenses that parallel those req uired by the means test for Chapter
7 debtors.51 To determine expenses that are not derived from IR S standards,
debtors will have to estimate future costs for items such as health insurance,
charitable contributions, and contributions to elderly , ill, or disabled family
members. Longitudinal research could track the accuracy of these estimates
and consider how differences between proj ected and actual expenses affect
the rate of plan completion.
A fter their Chapter 1 3 plan is confirmed, Chapter 1 3 debtors must file
annual statements at the req uest of the court, the U .S . trustee, or any party in
interest in the case as req uired by section 5 2 1 (f)(4 ). These annual statements
are different than the means test calculations prepared at the time of bankruptcy filing, and must report “income and expenditures of the debtor . . . that
[show] how income, expenditures, and monthly income are calculated” for
each y ear that the debtor is making Chapter 1 3 plan pay ments.52 A nationally
standardiz ed form for these annual reports has not been promulgated y et, and
it remains unclear how freq uently parties will req uest these annual statements. If widely available, these data on the income and expense traj ectories
of Chapter 1 3 debtors will illuminate the financial experiences of debtors
after filing bankruptcy . A precipitous drop in income as reported in an annual
statement may routinely precede the dismissal of a debtor’s Chapter 1 3 case.
A lternatively , perhaps debtors who report living on fixed incomes or entrepreneurs with income volatility are more or less likely to complete their
Chapter 1 3 plans. These theories may not sound controversial. The reality ,
49. I d . § 70 7(b) (2) (A) (ii) (I) (Supp. V 20 0 5) .
50 . Professor Eliz abeth Warren described BAPCPA’s procedural h urdles to
bankruptcy as “ a th ousand paper cuts.” See Paula Moore, C ritic s Say B ankruptc y
R ef o rm U nf airly Targets Small B usiness, DENV ER BU S. J. (Aug. 30 , 20 0 2) (q uoting
Professor Warren) .
51. Official Form B22C is available at th e U.S. Courts w ebpage, Bankruptcy
Forms, h ttp:/ / w w w .uscourts.gov/ bkforms/ index.h tml (last visited July 18, 20 0 6) .
52. 11 U.S.C. § 521(f) (4) (B) .
File: P o r t er ( f in a l) . d o c
9 7 2
C r ea t ed o n : 1 2 / 2 2 / 2 0 0 6 9 :0 5 :0 0 P M
M I S S O U R I L AW
R E V IE W
L a s t P r in t ed : 1 2 / 2 3 / 2 0 0 6 1 :1 1 :0 0 P M
[ V o l.7 1
however, is that the contours of Chapter 1 3 success are largely a my stery
because difficulties in compiling longitudinal data have frustrated research on
how debtors fare in Chapter 1 3 .53 The annual statements could spark a renewed debate about the efficacy of Chapter 1 3 .54
S imilar advances in understanding how small businesses fare in Chapter
1 1 may result from the new Chapter 1 1 reporting req uirements. S ection 3 0 8
req uires “small business debtor[s]”55 to “file periodic financial and other reports” that contain information on the debtor’s profitability , how the debtor’s
proj ected cash receipts and disbursements compare with actual figures, and
“such other matters as are in the best interests of the debtor and creditors, and
in the public interest in fair and efficient procedures.”56 A new M onthly O perating R eport form is apparently soon to be published for comment as a proposed official form.57 E nterprising researchers could compile data from these
reports to construct a sample of small business cases to examine how such
companies fare during bankruptcy . F or example, how q uickly do they become
profitable after filing? H ow accurate are debtor’s cash flow proj ections? This
window into the functioning of small business bankruptcy could support or
allay concerns that Chapter 1 1 is too expensive, slow, complicated or inefficient for small businesses.58
B . G ov ern m en t Research Stud i es
B A P CP A contains at least seven separate sections that instruct federal
public agencies to conduct studies of aspects of the bankruptcy sy stem. These
public studies draw a range of federal agencies into the universe of bankruptcy research. It seems doubtful that these studies will prove entirely fruitful, however, given the inherent problems with the scope of the studies to be
conducted and emerging problems with their execution.
53. Jean Brauch er, A n Empiric al Stud y o f D ebto r Ed uc atio n in B ankruptc y:
I mpac t o n C h apter 1 3 P lan C o mpletio n N o t Sh o w n, 9 AM. BANKR. INST. L. REV . 557
(20 0 5) .
54. I d . at 557 n.5 (collecting studies on Ch apter 13 plan completion) .
55. See 11 U.S.C. § 10 1(51D) (defining term) .
56. I d . § 30 8(b) .
57. U STP M aking “ G reat P ro gress” o n B A P C P A C h allenge, 46 BANKR. CT.
DEC . 6 (May 30 , 20 0 6) .
58. See, e.g., Douglas G. Baird & Edw ard R. Morrison, Serial Entrepreneurs and
Small B usiness B ankruptc ies, 10 5 COL U M. L. REV . 2310 (20 0 5) ; Ford Elsaesser, Th e
Small B usiness-B lues—M aking B ankruptc y W o rk in M id -Siz e C ases, AM. BANKR.
INST. J., WL 0 9180 3 ABI-CLE 547 (20 0 3) ; Robert M. Law less et al., A G limpse at
P ro f essio nal F ees and O th er D irec t C o sts in Small F irm B ankruptc ies, 1994 U. IL L .
L. REV . 847.
File: P o r t er ( f in a l) . d o c
2 0 0 6 ]
C r ea t ed o n : 1 2 / 2 2 / 2 0 0 6 9 :0 5 :0 0 P M
R E S E AR C H I N G B APC PA
L a s t P r in t ed : 1 2 / 2 3 / 2 0 0 6 1 :1 1 :0 0 P M
9 7 3
The G overnment A ccountability O ffice (“G A O ”) is the primary agency
responsible for study ing federal programs.59 A s noted previously , the G A O
has produced bankruptcy studies in the past and evaluated non-government
studies. S ection 2 0 5 of B A P CP A instructs the G A O to study the “reaffirmation agreement process . . . to determine the overall treatment of consumers
within the context of such process.”60 The study should consider “(1 ) the
policies and activities of creditors with respect to reaffirmation agreements;
and (2 ) whether consumers are fully , fairly , and consistently informed of their
rights pursuant to such title.”61
If taken seriously , this is a difficult charge. The activities of creditors
encompass behaviors that cannot be gleaned from court records. F or example,
the reaffirmation process includes private hallway conversations between
emotionally exhausted debtors and S ears’s representatives at the meetings of
creditors62 and the way in which attorney s counsel debtors on the wisdom of
reaffirming a particular debt. Crucial information about reaffirmation simply
is not q uantitative. A sensitive ethnographer or a skilled field interviewer
could perhaps illuminate these hidden processes, but the G A O report is
unlikely to do so.63 The G A O was also saddled with the weighty task of making “recommendations for legislation (if any ) to address any abusive or coercive tactics found in connection with the reaffirmation agreement process.”64
This ty pe of explicit policy assessment was performed by the bankruptcy
review commissions in previous rounds of bankruptcy law reform.65
The transfer of this ty pe of responsibility to the G A O could reflect several views, including that the N ational B ankruptcy R eview commissions were
too hostile to the credit industry or that the G A O is more speedy and efficient
than a commission. The G A O has contacted professors M arianne Culhane
and M ichaela W hite, authors of the leading academic study on reaffirmation,
59. Th e GAO describes its purpose on its h ome page. See Wh at is GOA? ,
h ttp:/ / w w w .gao.gov/ about/ w h at.h tml (last visited June 29, 20 0 6) .
60 . BAPCPA, Pub. L. No. 10 9-8, § 20 5, 119 Stat. 23, 50 (codified as amended at
11 U.S.C. § 363 (20 0 0 & Supp. V 20 0 5) .
61. I d .
62. 11 U.S.C. § 341 (20 0 0 & Supp. V 20 0 5) (req uiring debtors to attend a meeting w ith th eir creditors) . Th ese are often called “ 341 meetings” after th e sections of
th e Bankruptcy Code th at req uires th em.
63. Th e GAO is apparently conducting some interview s as part of th e study , but
it is not clear if th ese interview s are for background purposes to aid in study design or
if th ey w ill be part of th e data itself. Th e GAO w ill definitely rely on file data. Email
from Marianne Culh ane to K ath erine Porter (July 19, 20 0 6) (on file w ith auth or) .
64. BAPCPA § 20 5(b) .
65. Bankruptcy Reform Act of 1994, Pub. L. No. 10 3-394, 10 8 Stat. 410 6 (1994)
(current version codified as amended in scattered sections of 11 U.S.C.) . For more
details about th e ch arge given to th e most recent commission, see
h ttp:/ / govinfo.library .unt.edu/ nbrc/ facts.h tml.
File: P o r t er ( f in a l) . d o c
9 7 4
C r ea t ed o n : 1 2 / 2 2 / 2 0 0 6 9 :0 5 :0 0 P M
M I S S O U R I L AW
R E V IE W
L a s t P r in t ed : 1 2 / 2 3 / 2 0 0 6 1 :1 1 :0 0 P M
[ V o l.7 1
and asked for input on the study ’s design.66 This outreach is laudable. S erious
researchers build upon prior work, and the G A O should be commended for
taking this approach. The report was due in S eptember 2 0 0 6 ,67 but apparently
will be delay ed for an unknown period.68
B A P CP A req uires the G A O to conduct another study . Its purpose is to
evaluate the feasibility , effectiveness, and cost of req uiring trustees to report
the names and social security numbers of debtors to the O ffice of Child S upport E nforcement to determine whether debtors have outstanding support
obligations. A report was due “[n]ot later than 3 0 0 day s after the enactment of
[B A P CP A ].”69 B y my calculations, it should have been issued by F ebruary
2 0 0 6 , but it is not y et available.70 The delay illustrates the need to monitor the
implementation of the B A P CP A research provisions. It also raises some
doubt about Congress’s intent in req uesting such studies. R ather than reflecting a sincere desire for knowledge, the studies may be compromise legislation
that allowed congressional representatives opposed to B A P CP A to feel that
they will have some imprint on future bankruptcy policy making.
B A P CP A vests substantial new powers in the U S T program, including
implementing the debtor financial education program. B A P CP A req uires the
U S T to develop a pilot curriculum and to test the effectiveness of this program for eighteen months in six j udicial districts.71 A report is req uired that
evaluates the U S T’s financial education program as well as those “carried out
by the credit industry .”72 This report should complement the research of P rofessors G ross, B lock-Lieb, and W einer about debtor financial education and
will hopefully replicate some of their methodological techniq ues and incorporate their substantive insights.73 The U S T must also submit a report to the
congressional j udiciary committees that contains findings on the “impact”
66. Email from Marianne Culh ane to K ath erine Porter (July 19, 20 0 6, 20 :53
CST) (on file w ith auth or) .
67. Th e report w as not available at th e time th at th is volume w ent to press, but
sh ould be available at h ttp:/ / w w w .gao.gov.
68. Th e GAO h as stated th at it is “ currently determining th e scope of th e engagement and h a[ s] not establish ed our research obj ectives or w h en a report w ill be
issued” for th e reaffirmation study . Email from Orin B. Atw ater to K ath erine Porter
(July 20 , 20 0 6, 12:32 CST) (on file w ith auth or) .
69. BAPCPA § 230 (b) .
70 . In fact, correspondence from th e GAO suggests th at th e study is not even
underw ay . A GAO analy st reported th at th e GAO is “ in th e design ph ase of th e engagement and h ave not determined our research obj ectives or w h en a report w ill be
issued.” Email from Ellen T. Wolfe to K ath erine Porter (July 20 , 20 0 6, 10 :29 CST)
(on file w ith auth or) .
71. BAPCPA § 10 3.
72. I d . § 10 5.
73. Susan Block-Lieb, K aren Gross & Rich ard L. Wiener, L esso ns F ro m th e
Trenc h es: D ebto r Ed uc atio n in Th eo ry and P rac tic e, 7 FORDH AM J. CORP . & FIN.
L. 50 3 (20 0 2) ; Susan Block-Lieb & K aren Gross, D ebto r Ed uc atio n: M aking Sure a
G o o d I d ea D o es N o t G o A w ry, NORTON BANKR. L. ADV ISER 1, at 6 (Jan. 20 0 0 ) .
File: P o r t er ( f in a l) . d o c
2 0 0 6 ]
C r ea t ed o n : 1 2 / 2 2 / 2 0 0 6 9 :0 5 :0 0 P M
R E S E AR C H I N G B APC PA
L a s t P r in t ed : 1 2 / 2 3 / 2 0 0 6 1 :1 1 :0 0 P M
9 7 5
that using Internal R evenue S tandards to determine current monthly income
“has had on debtors and on the bankruptcy courts.”74 The U S T “may ” make
recommendations for amending the B ankruptcy Code based on its findings.
The point of this study , and what possible methodology could be adopted to
produce it, is ambiguous. W ith this broad of a charge, one could suspect that
the report will largely echo the overall position of the U S T supporting the
means test, rather than making specific data-driven observations. The U S T
may have been sensitive to such concerns, however, because it has outsourced the study to the R A N D Corporation.75
The U S T has announced its involvement in advising the A ttorney G eneral with regard to conducting the financial audits of individuals who file
bankruptcy .76 B eginning O ctober 2 0 , 2 0 0 6 , approximately 8 ,0 0 0 debtors will
be targeted for audits to determine the accuracy , veracity , and completeness
of their bankruptcy records.77 A report shall be produced for each audit and
filed with the court.78 It remains unclear whether researchers will be able to
conduct meaningful electronic searches of court records to locate these reports. The statute does req uire annual public disclosure of the aggregate results of these audits. H owever, the only statistic that is specifically req uired is
the “percentage of cases, by district, in which a material misstatement of income or expenditures is reported.”79 The audits will be expensive for taxpay ers,80 and a costly opportunity will be foregone if the audit data are not released in individual form that facilitates meaningful analy sis and evaluation
of the audit procedures. F or example, the audits should measure problems
with debtors overestimating the value of their assets, not j ust identify underestimations. This is also an issue of the “accuracy ” of schedules, one not considered by prior research.81 A nother benefit of audit analy sis could be the
identification of commonly omitted items and the development of software
74. BAPCPA § 10 3.
75. Email from Marianne Culh ane to K ath erine Porter (July 19, 20 0 6) (on file
w ith auth or) . Professor Culh ane and Wh ite w ill be consultants on th e study .
76. U STP M aking “ G reat P ro gress,” supra note 57.
77. I d .
78. BAPCPA § 60 3(a) (2) (A) .
79. I d . § 60 3(a) (2) (D) .
80 . Th e UST req uested $ 4.8 million for fiscal y ear 20 0 6 to fund th e audits. Testimo ny bef o re H . Subc o mm. o n C o mmerc ial and A d ministrativ e L aw , 10 9th Cong. 10
(Apr. 26, 20 0 6) (statement of Clifford J. Wh ite, Acting Director for th e Executive
Office
of
th e
UST)
av ailable
at
h ttp:/ / w w w .usdoj .gov/ ust/ eo/ public_ affairs/ testimony / docs/ testimony 0 60 426.pdf.
81. See Steven W. Rh odes, A P rev iew o f “ D emo nstrating a Serio us P ro blem
w ith U nd isc lo sed A ssets in C h apter 7 C ases,” 5 NORTON BANKR. L. ADV ISER 1 (May
20 0 2) (concluding th at debtors in asset cases underestimate asset values in some
situations) . Overvaluation could h arm debtors’ prospects of financial recovery because such estimations may be th e basis for reaffirmation negotiations or th e calculation of secured claims.
File: P o r t er ( f in a l) . d o c
9 7 6
C r ea t ed o n : 1 2 / 2 2 / 2 0 0 6 9 :0 5 :0 0 P M
M I S S O U R I L AW
R E V IE W
L a s t P r in t ed : 1 2 / 2 3 / 2 0 0 6 1 :1 1 :0 0 P M
[ V o l.7 1
tools and continuing education for debtor’s attorney s to help prompt clients to
initially disclose such assets.
B ey ond the G A O and the U S T, new bankruptcy legislation singled out a
few additional federal agencies to research bankruptcy issues.82 M ost notably ,
the S mall B usiness A dministration (“S B A ”) is req uired to study why small
businesses file bankruptcy .83 D ue in A pril 2 0 0 7 , this study is supposed to
identity way s in which “laws relating to bankruptcy may be made more effective and efficient in assisting small businesses to remain viable.”84 S ubstantial
scholarship has considered this q uestion already .85 Congress’s decision to
req uest a S B A study is subj ect to two interpretations. It may reflect distrust or
even disdain for academic research on bankruptcy , or it may be an indication
that Congress will ignore academic research unless a federal agency validates
those prior findings. Congress’s mandate to the S B A may also be understood
as a triumph for scholars. Their careful work may be responsible for prompting Congress to evaluate bankruptcy as a small business tool. E ither way , the
S B A study is noteworthy for forcing a government agency other than the
usual duo of the A O and U S T to acknowledge and evaluate the effects of
bankruptcy law on overall policy .
C . F uture Ap p roach to Stati sti cs
B A P CP A explicitly references the need for improved bankruptcy data.
A lthough only hortatory , section 6 0 4 expresses the “sense of Congress” that
the “national policy of the U nited S tates should be that all data held by bankruptcy clerks in electronic form . . . should be released in a usable electronic
form in bulk to the public.”86 The statute also seeks to improve the q uality of
bankruptcy statistics by stating that the bankruptcy data sy stem should use a
“single set of data definitions and forms . . . nationwide” and that “data for
any particular bankruptcy case [should be] aggregated in the same electronic
record.” E mpirical researchers should rej oice at these B A P CP A provisions,
which represent radical advances in existing data protocols. The improved
sy stem should reduce problems with data consistency and facilitate large or
national studies by reducing the cost of accessing records.87
82. See, e.g., BAPCPA § 130 8 (req uiring th e Federal Reserve Board to conduct a
study of th e bankruptcy impact of credit being extended to dependent college students) .
83. I d . § 443.
84. I d .
85. See Judge A. Th omas Small, I f Y o u F ix I t, Th ey W ill C o me — A N ew P laying
F ield f o r Small B usiness B ankruptc ies, 79 AM. BANKR. L.J. 981, 981 n.3 (20 0 5) (collecting articles on small business bankruptcy ) .
86. BAPCPA § 60 4.
87. V iew ed collectively , th e data provisions in BAPCPA represent at least a
partial victory for th e National Bankruptcy Review Commission and its Data Study
Proj ect. Some of th e statutes, such as section 60 2, are virtually identical to th ose sug-
File: P o r t er ( f in a l) . d o c
2 0 0 6 ]
C r ea t ed o n : 1 2 / 2 2 / 2 0 0 6 9 :0 5 :0 0 P M
R E S E AR C H I N G B APC PA
L a s t P r in t ed : 1 2 / 2 3 / 2 0 0 6 1 :1 1 :0 0 P M
9 7 7
B A P CP A has two distinct sections that directly address bankruptcy information. S ection 6 0 1 amends the U .S . Code statute on j udicial procedure to
add a section pertaining to “improved bankruptcy statistics.”88 S ection 6 0 2
adds req uirements pertaining to “bankruptcy data.”89 The A O is responsible
for implementing the former; the A ttorney G eneral is responsible for implementing the latter. The statutes are different in scope and purpose.
S ection 6 0 1 is concerned solely with debtors “who are individuals with
primarily consumer debts.”90 W hile the referenced statistics “shall” be made
available to the public, the statute’s main purpose is to facilitate annual reports to Congress. The annual reports will likely consist of aggregate statistics
about the macroeconomic impact of consumer bankruptcy and the demographics of consumer debtors. The list of req uired information is worth careful study and is too exhaustive to list here, but includes information on the
total assets, liabilities, income, and expenses of debtors.91 These reports will
reveal information that previously was only estimable through the painstaking
work of academic researchers such as the Consumer B ankruptcy P roj ects92
and the R eaffirmation P roj ect.93 W ith the new data, one should be able to
determine how many dollars of debt are discharged each y ear in consumer
cases and how debtors’ past and current incomes compare to average A mericans. F urther, studies about legal process will be possible and “local legal
culture” can be examined94 because the data are to be released “for each district.”95
The “bankruptcy data” provision, section 6 0 2 , applies to trustees’ final
reports in all cases and periodic reports in Chapter 1 1 cases. It req uires the
A ttorney G eneral, “within a reasonable time” after the enactment of
B A P CP A , to “issue rules req uiring uniform forms” for trustee’s reports.96 The
statute contains a laundry list of req uired information to be included in trustee
gested in th e Commission’s report on Data Compilation and Dissemination. See
BANKRU P TC Y , supra note 12, at 923-41; BAPCPA § 60 2.
88. BAPCPA § 60 1; 28 U.S.C. § 159 (Supp. V 20 0 5) .
89. BAPCPA § 60 2.
90 . I d . § 60 1(a) ; 28 U.S.C. § 159(a) .
91. BAPCPA § 60 1(c) (3) .
92. See TERESA A. SU L L IV AN, EL IZ AB ETH WARREN & JAY LAW RENC E WESTB ROOK, AS WE FORG IV E OU R DEB TORS 15-18 (1989) (describing data th e first Consumer Bankruptcy Proj ect collected th at th e government did not make available) .
93. See Culh ane & Wh ite, D ebt A f ter D isc h arge, supra note 6.
94. See generally Teresa A. Sullivan, Eliz abeth Warren & Jay Law rence Westbrook, Th e P ersistenc e o f L o c al L egal C ulture: Tw enty Y ears o f Ev id enc e f ro m th e
F ed eral B ankruptc y C o urts, 17 HARV . J.L. & PU B . POL ’Y 80 1 (1994) ; Jean Brauch er,
L aw yers and C o nsumer C ulture: O ne C o d e, M any C ultures, 67 AM. BANKR. L.J. 50 1
(1993) .
95. BAPCPA § 60 1(c) (2) .
96. I d . § 60 2(a) (codified as amended at 28 U.S.C. § 589b (Supp. V 20 0 5) .
File: P o r t er ( f in a l) . d o c
9 7 8
C r ea t ed o n : 1 2 / 2 2 / 2 0 0 6 9 :0 5 :0 0 P M
M I S S O U R I L AW
L a s t P r in t ed : 1 2 / 2 3 / 2 0 0 6 1 :1 1 :0 0 P M
R E V IE W
[ V o l.7 1
reports.97 The standardiz ation and disclosure of these reports was recommended by the N ational B ankruptcy R eview Commission report in 1 9 9 4 and
largely represent the fulfillment of the Commission’s recommendations in
this regard.98
The unknown factor about the new data req uirements is the extent to
which the government will make individualiz ed data available in a useful
format. S ome statutes are unclear on what exactly must be released. S ection
6 0 1 ’s reference to “statistics” rather than data arguably means that only aggregated figures must be publicly disclosed. S ection 6 0 2 explicitly req uires
that trustees’ reports should be filed “so as to facilitate compilation of data
and maximum possible access of the public.”99 H owever, this req uirement
could be satisfied merely by making the trustee reports available on the existing P A CE R court record sy stem. S ection 6 0 4 suggests a broader reading of
public access is req uired by opining that national policy should be that all
public record data “be released in a usable electronic form in bulk to the public.”
The degree to which the government effectuates this “sense of Congress” provision will be crucial in determining if B A P CP A facilitates empirical research. W ithout individualiz ed data in formats accessible to statistical
software, research will be sharply curbed. The reference to the “public” in
S ection 6 0 4 is also intriguing. This should encompass not j ust the occasional
curious citiz en and academic scholars, but also for-profit researchers and the
credit industry . B anks or lenders could harness these bankruptcy data to
change their litigation tactics in bankruptcy cases or improve predictions
about bankruptcy charge-offs. If all parties have eq ual access to raw data,
academic researchers must continue to conduct empirical research or risk the
domination of bankruptcy research by for-profit entities which have dollars at
stake in the research results.10 0
IV . PITFALLS IN RESEARCH AFTER BAPCPA
A. Data G ap s
The potential of B A P CP A to improve bankruptcy research is vast. The
research provisions arguably represent congressional acknowledgement that
documenting the realities of the bankruptcy sy stem — who uses it, what
problems they bring to it, and how it functions — are critical to crafting effective bankruptcy policy . A dditionally , the amended statutes themselves
should force the disclosure of heretofore hidden data. D espite these advances,
pitfalls remain for future bankruptcy empiricists.
97.
98.
99.
10 0 .
BAPCPA § 60 2(d) , (e) .
See BANKRU P TC Y , supra note 12, at 934.
BAPCPA § 60 2(b) .
See Warren, Th e M arket f o r D ata, supra note 6, at 7-8.
File: P o r t er ( f in a l) . d o c
2 0 0 6 ]
C r ea t ed o n : 1 2 / 2 2 / 2 0 0 6 9 :0 5 :0 0 P M
R E S E AR C H I N G B APC PA
L a s t P r in t ed : 1 2 / 2 3 / 2 0 0 6 1 :1 1 :0 0 P M
9 7 9
The usefulness of much B A P CP A data will be limited because of a lack
of comparative data. Initially , it will be nearly impossible to draw any inferences from the new statistics about the changes made by B A P CP A because
comparable figures do not exist for the pre-B A P CP A period. A lthough this
problem will dissipate, allowing for longitudinal comparisons after B A P CP A ,
we simply will never have much of this information for the y ears before
2 0 0 5 . B A P CP A does nothing to resolve a parallel research problem, which is
the lack of comparable figures for the general (non-bankrupt) A merican
population. The U .S . Census and the S urvey of Consumer F inances collect
only limited ty pes of financial information, and the methods used in these
studies do not parallel the bankruptcy process. The result is that often the data
are not comparable. The more serious difficulty arises when try ing to compare debtors to others in financial distress that have not filed bankruptcy .10 1
Identify ing these individuals is hard or impossible,10 2 as is garnering research
participation from them.10 3 M any believe that B A P CP A creates barriers to
filing bankruptcy , which will discourage or prevent some people from seeking bankruptcy relief.10 4 Y et, it will be challenging to document this trend and
to study those excluded from the sy stem. R ecent improvements in state court
records offer some data on foreclosures, j udgments, and garnishments that
could help identify families in financial trouble who may not file bank-
10 1. Th e same problem exists in th e business bankruptcy context but is less troubling. A forth coming paper by Professor Edw ard Morrison, Selec ting B ankruptc y: A n
Empiric al Stud y o f Small B usiness D ebto r (manuscript on file w ith auth or) uses Dun
& Bradstreet data to examine th e traj ectories of companies in financial trouble th at do
not file bankruptcy , comparing th em to th ose th at do file bankruptcy .
10 2. See RONAL D MANN, CH ARG ING AH EAD: TH E GROW TH AND REG U L ATION OF
PAY MENT CARD MARKETS AROU ND TH E WORL D 186 (20 0 6) (noting difficulty in collecting data on people w h o do not w h ile bankruptcy but are in comparable financial
circumstances to th ose w h o file bankruptcy ) .
10 3. Social scientists h ave found th at people find talking about money almost as
difficult as talking about sex. See Eleanor Singer, P ublic R eac tio ns to So me Eth ic al
I ssues o f So c ial R esearc h , 11 J. CONSU MER RES. 50 1, 50 4 (June 1984) (noting th at sex
and income w ere most freq uently mentioned topics th at offended research participants, beating out mental h ealth , drug use, and religion by w ide margins) .
10 4. See Ronald J. Mann, B ankruptc y R ef o rm and th e “ Sw eat B o x ” o f C red it C ard
D ebt,
U.
IL L .
L.
REV .
(forth coming
20 0 6) ,
av ailable
at
h ttp:/ / papers.ssrn.com/ sol3/ papers.cfm? abstract_ id= 89540 8 (arguing th at th e purpose
of consumer bankruptcy reform w as to delay filing, th ereby driving up th e amount of
fees and interest ow ed by consumers in financial trouble) ; Jean Brauch er, M eans
Testing C o nsumer B ankruptc y: Th e P ro blem o f M eans, 7 FORDH AM J. CORP . & FIN. L.
40 7, 411 (20 0 2) (“ All th ese new burdens w ould make Ch apter 7 more cumbersome
and th us more expensive and inaccessible.” ) ; Tabb, D eath o f C o nsumer B ankruptc y,
supra note 1, at 34 (“ Th e Reform Bill contains additional provisions th at also w ill
create entry barriers [ to bankruptcy .]” ) .
File: P o r t er ( f in a l) . d o c
9 8 0
C r ea t ed o n : 1 2 / 2 2 / 2 0 0 6 9 :0 5 :0 0 P M
M I S S O U R I L AW
R E V IE W
L a s t P r in t ed : 1 2 / 2 3 / 2 0 0 6 1 :1 1 :0 0 P M
[ V o l.7 1
ruptcy .10 5 A gain, however, a lack of prior state court research will limit inferences about how B A P CP A has affected the consumer bankruptcy filing decision.
B . Research O b stacl es
B A P CP A increases the roles of the A O and the U S T as custodians of
bankruptcy data and as researchers responsible for generating bankruptcy
statistics. H owever, heightened government participation in data collection is
not without peril. These agencies must balance debtors’ privacy , the public’s
desire for full access, and efficient functioning of the bankruptcy sy stem. The
A O ’s prior actions with regard to bankruptcy data suggest an institutional
reluctance to share data and a failure to view responsibility for data as essential to the J udicial Conference’s mission. A similar statement could be fairly
made about the U S T. F or example, some bankruptcy court clerks argued that
section 1 0 7 of the B ankruptcy Code, which makes papers filed in bankruptcy
cases “public records . . . open to examination”10 6 did not extend to data in
electronic form. A similarly pinched reading of section 6 0 4 could be wielded
by local or federal bankruptcy officials. A lso, a collision between statistics
and politics is certainly plausible, as the top officials at the U S T and A O are
appointed by the P resident and the Chief J ustice of the S upreme Court, respectively , and could come under pressure to control the use of government
data. D espite B A P CP A ’s “sense of Congress” provision about public access
to data, there remains little or no recourse for researchers who are denied
access to data.
The U S T’s new supervisory responsibilities under B A P CP A , such as
approving credit counselors and financial education programs, may increase
its influence on releasing data. The need for U S T approval may discourage
these counseling and education entities from participating in controversial
academic studies for fear that cooperation could displease the U S T. M ore
alarmingly , rumors have circulated that the U S T may have considered banning researchers from conducting studies at meetings of creditors,10 7 but thus
far this has not occurred. The U S T could also effectively prohibit or condition
research by directing panel trustees, who are chosen by the U S T, to refuse to
cooperate in any studies.10 8 S uch research protocols could eq ually harm credit
industry and academic researchers, both of whom have relied on panel trustees and meetings of creditors to execute studies. E ven worse, the A O or U S T
10 5. For example, Wisconsin makes data from each of its county courts available
electronically and includes a data extraction option for a subscription fee. See Wisconsin Circuit Court Access, h ttp:/ / w cca.w icourts.gov/ index.xsl (last visited Oct. 11,
20 0 6) .
10 6. 11 U.S.C. § 10 7(a) (Supp. V 20 0 5) .
10 7. Each debtor is req uired to attend a meeting of creditors under 11 U.S.C. §
341(a) (20 0 0 ) . Th ese meetings are open to th e public.
10 8. See 28 U.S.C. § 586(a) (1) (20 0 0 ) ; 28 CFR § 58.1 (20 0 6) .
File: P o r t er ( f in a l) . d o c
2 0 0 6 ]
C r ea t ed o n : 1 2 / 2 2 / 2 0 0 6 9 :0 5 :0 0 P M
R E S E AR C H I N G B APC PA
L a s t P r in t ed : 1 2 / 2 3 / 2 0 0 6 1 :1 1 :0 0 P M
9 8 1
could “play favorites” in choosing when and to whom to release data.
B A P CP A contains no obvious remedy for this ty pe of grievance.
B A P CP A did not address one of the most sensitive issues facing empirical legal researchers. The federal government’s sy stem for obtaining court
documents, P A CE R , currently charges a fee of eight cents per page
viewed.10 9 F or large empirical proj ects, fees can q uickly total thousands of
dollars. S cholars may obtain a fee waiver by obtaining the written agreement
of the chief j udge of the district in which they wish to view records.110 N ot
only is this procedure cumbersome for national studies, but it also raises the
specter of censorship. J udges may be reluctant to facilitate studies that could
be used to criticiz e the operation of the bankruptcy sy stem in their j urisdictions. The j udiciary ’s outrage following the publication of Ly nn LoP ucki’s
book, C o urting F ailure ,111 has only exacerbated this problem, or at least perceptions of it.112 O ne solution would be to appoint a board consisting of bankruptcy j udges, government officials from the A O , U S T, G A O or other agencies, bankruptcy attorney s, and academic researchers to review proposals for
fee waivers. The review should not be substantive; the panel’s sole purpose
should be to verify that the req uested data will be used exclusively for nonprofit research.113
A s the debate over the $ 4 0 0 “bankruptcy tax” illustrated,114 the credit
industry uses data to advocate for their desired bankruptcy policies. Lenders
have a substantial advantage over academic scholars in generating research,
aside from the obvious fact that the industry has profit dollars at stake that
j ustify large research budgets. Credit transactions are almost alway s private
and conducted solely between the parties. B anks claim that much critical data
about the consumer credit sy stem is proprietary and that they currently are
10 9. See Public Access to Court Electronic Records: Overview ,
h ttp:/ / pacer.psc.uscourts.gov/ pacerdesc.h tml (last visited Nov. 17, 20 0 6) .
110 . See
Electronic
Public
Access
Fee
Sch edule,
h ttp:/ / pacer.psc.uscourts.gov/ documents/ epa_ feesch ed.pdf (last visited Nov. 17, 20 0 6)
(explaining fee exemption policy and stating th at exemption may be revoked at any
time at discretion of court) .
111. LY NN M. LOPU C KI, COU RTING FAIL U RE (20 0 5) (using empirical data to assert
th at competition betw een bankruptcy courts for large cases is producing undesirable
and inefficient case outcomes) .
112. See Brent Snavely , L o c al J ud ge P ush es to M ake F irms F ile f o r B ankruptc y
o n H o me Turf , CRAIN’S DETROIT BU S., May 16, 20 0 5, at 43 (q uoting Judge Steph en
Rh odes w h o ch aracteriz ed th e book as “ an inflammatory attack” ) .
113. Th e National Bankruptcy Review Commission called for th e appointment of
a “ Bankruptcy Data Coordinator” to oversee data collection betw een th e AO and th e
UST. See BANKRU P TC Y , supra note 12, at 921. Wh ile BAPCPA did not appoint such
a person, th is could be an alternate meth od for coordinating data release and creating
uniform and unbiased policies.
114. See Warren, Th e M arket f o r D ata, supra note 6, at 14 (describing th e development of th e $ 40 0 bankruptcy cost figure and h ow it sh aped debate over bankruptcy
reform) .
File: P o r t er ( f in a l) . d o c
9 8 2
C r ea t ed o n : 1 2 / 2 2 / 2 0 0 6 9 :0 5 :0 0 P M
M I S S O U R I L AW
R E V IE W
L a s t P r in t ed : 1 2 / 2 3 / 2 0 0 6 1 :1 1 :0 0 P M
[ V o l.7 1
under no obligation to share most of this data.115 B anks know, for example,
what percentage of dollars that consumers charge to credit cards is paid to
hospitals, doctors, or medical facilities. A cademics researching the medical
causes of bankruptcy are disadvantaged without these data, y et private industry can produce such figures at their whim to interj ect into policy debates.
S cholars cannot attack these numbers because the underly ing proprietary data
remains protected from scrutiny . B y creating a larger universe of publicly generated data, B A P CP A may alleviate some of this tension between academic and industry research, but will not entirely alleviate the problem.
B A P CP A coincidentally generates new proprietary data by req uiring potential debtors to go through credit counseling and financial education. These
organiz ations will have useful information about who considers bankruptcy
from credit counseling and about the post-filing circumstances of families
from financial education sessions. A consumer advocacy group has already
released a study using credit counseling data gathered from willing agencies.116 F uture partnerships with these entities could produce other fruitful
collaborations, although most of the credit industry has generally refused
access to legal scholars. This relationship contrasts with the incredible access
that banks and other corporations grant to business school professors seeking
to write case studies or conduct marketing or finance research. Law has
adopted a more independent research model, free from actual or implied bias
by contact with private industry . The drawback to this approach is that problems with cost and access more freq uently limit the scope of legal research.
V . CONCLUSION
B A P CP A made sweeping changes to the B ankruptcy Code as written.
This sy mposium highlights some key elements of these reforms and speculates on how B A P CP A will function. The challenge going forward is for researchers to explore the realities of the amended B ankruptcy Code as experienced. B A P CP A provides tremendous opportunities to obtain new data and
develop fresh approaches for empirical research. The time to attack potential
problems with bankruptcy data is now, and the task is to ensure that these
data and studies are valid, accessible, and meaningful. To capitaliz e on these
opportunities, the federal government, scholars, attorney s, j udges, and industry should cooperate in the design and execution of this research and should
critically analy z e what available data reveal about the bankruptcy sy stem.
S cholars can and should contact respective government agencies to offer their
expertise in designing empirical studies. They should req uest in advance that
the data be released in way s that will be most useful for analy sis, rather than
115. Th e Home Mortgage Disclosure Act does req uire th e disclosure of information about h ome loans, see 12 U.S.C. § § 280 1-2810 (20 0 0 ) , but th ere is no similar
req uirement for most oth er ty pes of consumer debt.
116. DEADB EATS, supra note 2.
File: P o r t er ( f in a l) . d o c
2 0 0 6 ]
C r ea t ed o n : 1 2 / 2 2 / 2 0 0 6 9 :0 5 :0 0 P M
R E S E AR C H I N G B APC PA
L a s t P r in t ed : 1 2 / 2 3 / 2 0 0 6 1 :1 1 :0 0 P M
9 8 3
lamenting these problems after the fact, and should offer concrete models
such as the U .S . Census B ureau’s A merican F actfinder service117 and the
D epartment of J ustice’s B ureau of J ustice S tatistics to the A O and U S T as
examples of how to release data to the public.118 These agencies should be
forthcoming with information about pending studies and invite public comment on their proposed methodologies. A t this point, the predominant model
seems to be obscurity and privacy . If the B A P CP A studies are produced in
isolation and the release of new data is limited, the credibility of the federal
government as a purvey or of bankruptcy research will be damaged. Transparent and cooperative data collection could eliminate tensions between academic and industry researchers and provide a foundation for a new level of
bankruptcy knowledge. B A P CP A ’s potential to improve bankruptcy data is
vast. A voiding the perils of these new studies and data collection policies
req uires timely and sustained efforts from all parties concerned about accurate, timely , and reliable bankruptcy data.
117. Th e Census Bureau makes detailed data available to be dow nloaded and
imported into leading statistical programs. See U.S. Census Bureau American FactFinder Home Page, h ttp:/ / factfinder.census.gov/ h ome/ saff/ main.h tml? _ lang= en (last
visited Nov. 17, 20 0 6) .
118. Th e Bureau of Justice Statistics is a component of th e Department of Justice
and h as existed for tw enty -five y ears. See Bureau of Justice Statistics Home Page,
h ttp:/ / w w w .oj p.usdoj .gov/ bj s (last visited Nov. 17, 20 0 6) .